This article concentrates on the problem of practically time-synchronized tracking control for multi-input multi-output (MIMO) systems with unmatched nonlinearities and input saturations. Different from the existing approaches, a practically time-synchronized command filtered backstepping (CFB) control scheme is proposed. By integrating modified command filters and control signals designed with norm-normalized sign functions, the newly developed framework not only retains the advantages of the CFB control approach but also guarantees the property of time-synchronized convergence. Specifically, the "explosion of complexity" phenomenon and the influence of filtering errors are simultaneously addressed, and all components of the tracking error can achieve practically synchronous convergence to a small neighborhood of the origin in a finite time, despite the presence of unmatched nonlinearities in high-order systems. Furthermore, novel auxiliary systems are recursively embedded into each step of the time-synchronized CFB design to counteract the effect of input saturation. Rigorous theoretical analyses and comparative simulations demonstrate the rationality, effectiveness, and superiority of the proposed control scheme.

Effects of gas recirculation on the performance of solid oxide fuel cell-based auxiliary power system (SOFC-APS)

Abstract In order to address the complex faults and immature handling methods associated with the diverse equipment in the insulation board production line, a fault knowledge graph construction method based on a BERT-BiGRU-CRF model was proposed. Through analyzing fault knowledge characteristics and defining ontologies with the seven-step method, entity labels are calibrated, relationship rules established, and datasets formed according to mode layer Settings. By integrating Bidirectional Encoder Representations from Transformers (BERT), Bidirectional Gated Recurrent Unit (BiGRU), and Conditional Random Fields (CRF), the faulty entities are extracted, and entities with similar textual expressions are aligned through the proposed entity alignment method based on weighted average similarity. On our dataset, the model achieved an optimal precision (0.9012) and an F1 score (0.9107), demonstrating outstanding performance in the tasks of fault entity identification. Moreover, the constructed fault knowledge graph, stored as triples in Neo4j database, serves as the foundation for an auxiliary diagnostic system designed to deliver precise and reliable fault diagnosis support.

Multi-sensor fusion-based intelligent auxiliary system of power wheelchairs for individuals with limbs disabilities: design and implementation

Nowadays, 2 to 4 inverters are installed in a metro train to feed the non-traction electrical loads (auxiliary loads). Increasingly, the reliability of this Auxiliary power supply systems (APSS) is paramount, ensuring uninterrupted operation of vital subsystems such as lighting, ventilation. In this paper, a novel approach to enhance the reliability and performance of APSS is presented. Following an introduction to the conventional APSS structure, a new framework is proposed through four key steps: 1) Data input, 2) Revamped classification of auxiliary loads based on technical parameters, 3) Strategic allocation of power sources to load groups, and 4) Enhancement of reliability indices. Through a detailed case study (Tehran metro’s trains), the superiority of the new APSS is demonstrated, showcasing improved control features, heightened reliability, enhanced fault tolerance, and reduced implementation costs. The analysis and results reveals a significant improvement in reliability, with the new APSS exhibiting a maximum non-reliability rate of 15.4%, compared to 26.5% for the conventional system. Furthermore, the findings highlight the new APSS's ability to maintain power supply to high-priority load groups even after multiple faults. Finally, cost analysis indicates a more economical implementation of the new APSS, with lower equipment installation and cabling costs.

This paper investigates a fixed-time fault-tolerant control problem for the virtually coupled train set subject to the actuator faults, input saturation, external disturbances and parameter uncertainties. A novel fixed-time adaptive sliding mode fault-tolerant control scheme is proposed. By introducing the hyperbolic tangent function, a failure compensation strategy is designed to address the unknown actuator fault without the prior knowledge of fault. Moreover, we estimate the combinations of unknown parameters instead of directly estimating the unknown parameter itself, which significantly saves the computing resources of trains and improves the efficiency. Furthermore, an auxiliary system is introduced to compensate for the influence of input saturation. It guarantees the position and speed tracking errors of each train can converge to the small regions in fixed time in despite of the influence of internal faults and external disturbances, as well as actuator saturation. The proposed control scheme is of great significance to improve the efficiency, safety and reliability of high-speed trains' operation. Finally, simulations are provided to verify the feasibility and effectiveness of the proposed control scheme.

To further enhance economic efficiency and optimize energy conservation and emission reduction performance, an optimized energy management strategy (EMS) tailored for the hybrid power system of rubber-tyred gantry cranes is proposed. Wavelet packet decomposition (WPD) was employed as the signal processing approach, and this method was further integrated with EMS for hybrid power systems. Through a three-layer progressive architecture comprising WPD frequency–domain decoupling, fuzzy logic real-time adjustment, and PSO offline global optimization, a cooperative optimization mechanism has been established in this study between the frequency-domain characteristics of signals, the physical properties of energy storage components, and the real-time and long-term states of the system. Firstly, the modeling and simulation of the power system were conducted. Subsequently, an EMS based on WPD and limit protection was developed: the load power curve was decomposed into different frequency bands, and power allocation was implemented via the WPD algorithm. Meanwhile, the operating states of lithium batteries and supercapacitors were adjusted in combination with state of charge limits. Simulation results show that this strategy can achieved reasonable allocation of load power, effectively suppressed power fluctuations of the auxiliary power unit system, and enhanced the stability and economy of the hybrid power system. Afterward, a fuzzy controller was designed to re-allocate the power of the hybrid energy storage system (HESS), with energy efficiency and battery durability set as optimization indicators. Furthermore, particle swarm optimization algorithms were adopted to optimize the EMS. The simulation results indicate that the optimized EMS enabled more reasonable power allocation of the HESS, accompanied by better economic performance and control effects. The proposed EMS demonstrated unique system-level advantages in enhancing energy efficiency, extending battery lifespan, and reducing the whole-life cycle cost.

Turbocharged diesel engines are widely used for the propulsion and as the generators for powering auxiliary systems in marine applications. Many works were published on the development of diagnosis tools for the engines using data from simulation models or from experiments on a sophisticated engine test bench. However, the simulation data varies a lot with actual operational data, and the available sensor data on the actual vessel is much less compared to the data from test benches. Therefore, it is necessary to develop anomaly prediction and fault diagnosis models from limited data available from the engines. In this paper, an artificial intelligence (AI)-based anomaly detection model and machine learning (ML)-based fault diagnosis model were developed using the actual data acquired from a diesel engine of a cargo vessel. Unlike the previous works, the study uses operational, thermodynamic, and vibration data for the anomaly detection and fault diagnosis. The paper provides the overall architecture of the proposed predictive maintenance system including details on the sensorization of assets, data acquisition, edge computation, and AI model for anomaly prediction and ML algorithm for fault diagnosis. Faults with varying severity levels were induced in the subcomponents of the engine to validate the accuracy of the anomaly detection and fault diagnosis models. The unsupervised stacked autoencoder AI model predicts the engine anomalies with 87.6% accuracy. The balanced accuracy of supervised fault diagnosis model using Support Vector Machine algorithm is 99.7%. The proposed models are vital in marching towards sustainable shipping and have potential to deploy across various applications.

ABSTRACT This paper presents an event‐triggered distributed control protocol designed for scenarios involving Denial‐of‐Service (DoS) attacks within directed communication graphs. We address the consensus problem of uncertain systems by employing neural networks to approximate nonlinear components. Compared with traditional neural network observers, the proposed method incorporates damping terms to ensure the boundedness of neural network estimators. Initially, an auxiliary system was constructed to decouple the consensus problem under a directed communication graph framework. It is demonstrated that the uncertain nonlinear system can achieve consensus through the utilization of information exchanged among neighboring agents. Subsequently, to curtail the consumption of communication resources, we introduce an event‐triggered control protocol based on the aforementioned approach, incorporating an effective event‐triggering function. Building upon the proposed control protocol, we further develop a control strategy tailored for systems under DoS attacks. Furthermore, the event‐triggered mechanism without using global information in trigger condition not only curtails superfluous information exchange among agents but also precludes the occurrence of Zeno behavior (i.e., infinite triggering within a finite time interval). Finally, numerical simulations are carried out to validate the efficacy of the proposed control protocol.

Target tracking, a fundamental problem in multi-agent networks, has been widely studied in recent years. The communication channels in agents are vulnerable to external eavesdroppers. Thus, how to defend against eavesdroppers has attracted a great attention in multi-agent networks. In this paper, a secure target tracking algorithm is proposed in the case that there are eavesdropping and random packet losses during the communication process. The agents estimate the state of target by performing collaborative state estimation based on average consensus Kalman filtering at every time slot and then update their own states to achieve target tracking. The agents in the network transmit state information of their auxiliary systems to defend against eavesdroppers. A novel noise generator is designed based on a neural network, and the artificial noise is injected into the information to be transmitted, such that each agent can exactly know the noise injected into the information received from its in-neighbour, while the eavesdropper cannot estimate the noise. The parameter condition of the auxiliary system guarantees secure target tracking against the eavesdropper. A simulation example is given to illustrate the effectiveness of the proposed algorithm.

Abstract Investigation of dual-frequency (DF) effects in helicon plasmas were conducted on the PKU Plasma Test (PPT) device, integrating a 13.56 MHz helicon source with a 2 MHz auxiliary heating system under cylindrical and magnetic mirror configurations. Experimental results demonstrate that lower-frequency (LF, 2 MHz) auxiliary heating effectively elevates electron temperature (T_e) while reducing ion density (n_i), thereby overcoming the limitation of single-frequency (SF) helicon systems in achieving efficient electron heating. The observed density reduction is attributed to the accelerated axial flow velocity, validated via 10-tip Mach probe diagnostics, and suppressed radial particle transport confirmed by 5-tip Langmuir probe array measurements. DF operation of helicon plasma in magnetic mirror field shows similar T_e elevation and lower density reduction amplitude. Moreover, the linear response of T_e to LF power offers a viable strategy for temperature adjustment.

This paper addresses the trajectory tracking problem for a six-degree-of-freedom (6-DOF) underwater manipulator subject to complex disturbances and input saturation. It proposes a fixed-time preset performance sliding mode control method considering input saturation (FT-PP-SMC-IS), aiming to achieve rapid and stable tracking performance under these constraints. Firstly, to improve modeling accuracy, the Newton–Euler method and Morison’s equation are integrated to establish a more precise dynamic model of the underwater manipulator. Secondly, to balance dynamic and steady-state performance, a preset performance function is designed to constrain the tracking error boundaries. Based on dual-limit homogeneous theory, a fixed-time sliding mode surface is constructed, significantly enhancing the convergence speed and fixed-time stability. Furthermore, to suppress the effects of input saturation, a fixed-time auxiliary system is designed to compensate in real-time for deviations caused by actuator saturation. By separately constructing the sliding mode reaching law and equivalent control law, global fixed-time convergence of the system states is ensured. Based on Lyapunov stability theory, the fixed-time stability of the closed-loop system is rigorously proven. Finally, comparative simulation experiments verify the effectiveness and superiority of the proposed method.

Evanescent field illumination (EFI) offers a promising solution to overcome the fundamental limitations in illumination uniformity, signal-to-noise ratio (SNR), and detection throughput for quantitative fluorescence analysis. The development of EFI chips faces tradeoffs between system complexity, illumination area, and manufacturing scalability. In this paper, a compact optical waveguide chip capable of generating excellent detection throughput (large-scale), highly uniform EFI via total internal reflection (TIR) on a waveguide surface is presented that consists solely of an integrated aspheric coupling lens and a standard microscope slide-sized planar waveguide (75 mm × 25 mm). High-precision (Ra < 5 nm), mass-producible fabrication of this chip using polystyrene (PS) through injection molding processes is demonstrated. Experimental results reveal a substantial 2.01 mm alignment tolerance in light coupling while producing highly uniform EFI over large areas (>12 mm × 12 mm, CV < 6%), requiring no complex auxiliary systems.

With the gradual development of smart power plants and large-scale centralized control, there is a need to exchange a large number of signals between different DCS systems and between DCS and PLC systems. Different control systems have different brands and cannot communicate directly via networks. Moreover, due to network security concerns, the main control of unit units and the auxiliary control system of the entire plant cannot communicate directly via networks either. The commonly adopted methods for signal exchange between control systems are hardwiring and 485 communications. Both have obvious drawbacks, where hardwiring requires a large number of channels and cable laying; 485 configuration is difficult, not easy to maintain, and faults are hard to locate. This paper studies how to strike a balance between the two, using a minimal amount of hardwiring to transmit a large number of signals, which is safe, reliable, cost-effective, and can be maintained by any control personnel without network security risks.

AimsThis study aims to overcome the limitation of integrated AI in telemedicine by developing MedAssist, an end-to-end intelligent auxiliary system designed to leverage multimodal patient data across the entire care continuum.MethodsWe built MedAssist upon a novel lightweight multimodal medical foundation model, establishing a standardized data platform and a specialized model architecture to fuse textual and visual data.1 The model was compressed using knowledge distillation and quantization-aware training to enable rapid deployment while maintaining high accuracy.ResultsEvaluated on a multi-center dataset, the model achieved an AUC of 0.87 in a critical diagnostic task and reduced inference latency by 60%. A pilot clinical validation demonstrated a 40% reduction in triage time and a 92% concordance rate with senior physician diagnoses.ConclusionThe findings indicate that MedAssist provides a robust and scalable framework for enhancing the efficiency and quality of remote healthcare, representing a significant step towards practical, AI-augmented telemedicine.

This study investigates the combined effect of anodic pretreatment and varying auxiliary formulations on the dyeing performance of cotton fabric using a natural dye extracted from Butea monosperma flowers. While past studies have examined the role of mordants and pH modifiers in natural dyeing, limited attention has been given to how auxiliary systems interact with electrochemically modified fabrics to influence colour yield and fastness performance. This research addresses that gap by evaluating how anodically pretreated cotton responds to common dyeing auxiliaries—sodium chloride (NaCl), potassium aluminium sulfate (alum), sodium carbonate $({\rm Na_2 CO_3}),$ and calcium oxide (CaO)—in terms of shade development, chromatic richness, and durability. Using the CIE Lab colour system, significant colour variations were observed across all treated specimens (2–7) compared to the salt-only control (specimen 1). ∆E values ranged from 12.49 to 19.95, with the highest chromatic shift seen in the ${\rm NaCl}^+$ alum sample, which produced a vivid orange hue with enhanced redness and yellowness. Combinations of alum and alkalis yielded deeper, more saturated shades, while alkali-only formulations produced earthier tones with moderate lightness. These outcomes highlight the synergistic effect of anodic surface activation and auxiliary chemistry on improving dye-fibre interactions. Fastness tests, conducted according to ISO 105-C10 and ISO 105-X12 standards, revealed moderate washing durability (colour change ratings of 3 to 4) and good rubbing fastness (ratings up to 4). Specimens treated with mordants and alkalis generally exhibited improved fixation and reduced staining, whereas anodic treatment enhanced surface adhesion of the dye. This study demonstrates a promising, low-impact approach to improving natural dye uptake and fastness on cotton fabric by integrating anodic pretreatment with accessible auxiliary systems. The findings support the development of sustainable, decentralised dyeing processes using locally available plant waste, with particular relevance for rural craft and textile communities seeking eco-friendly alternatives to synthetic dyes.

Background: In recent years, major diseases such as liver cancer have magnified severe threats to public health, with low early detection rates and inadequate real-time evaluation of treatment effects. Conventional ultrasound diagnosis relies heavily on physicians’ subjective experience, leading to high rates of missed or misdiagnosed cases, and intelligent auxiliary diagnosis systems lack dynamic evaluation capabilities for the full clinical diagnosis. Methods: To solve these issues, this study develops an intelligent medical auxiliary platform based on Scatterer Spacing (SS) estimation. There are many functions including ultrasound RF echo signal and B-mode image acquisition, time-frequency domain analysis, SS quantitative estimation. Results: The experimental results show that the LOMGA-based mean ISS decreased by 35.00% compared with 22.83% for WT, as well as the Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC) curve for distinguishing normal and coagulated tissues reaches 0.97, compared with 0.88 for WT. Conclusion: This platform reduces reliance on physicians’ subjective experience, improves the accuracy of early disease diagnosis and real-time evaluation of treatment efficacy, and promotes the clinical translation of quantitative ultrasound tissue characterization technology.

ABSTRACT In this article, a fixed‐time event‐triggered fuzzy adaptive control strategy is investigated for underactuated unmanned surface vehicles (USVs) formation. Initially, the underactuated issue is tackled by transforming the mathematical model of USVs. Then, the preset formation is achieved by the leader‐follower approach, and a fixed‐time differentiator is employed to obtain the real‐time differential signals of the virtual control law, which reduces the complexity of the controller design. Fuzzy logic systems (FLSs) are employed to approximate model uncertainties and external disturbances, and fixed‐time convergence is obtained. Furthermore, a fixed‐time relative threshold event‐triggered controller is designed based on an asymmetric differentiable saturation function, and an auxiliary saturation system is introduced to attenuate the effect of input saturation. Finally, the boundedness of the signal in the closed‐loop system is proved by the Lyapunov theory. Simulation experimental results verify the effectiveness of the proposed strategy.

Relaxed quantum systems with conservation laws are believed to be approximated by the generalized Gibbs ensemble (GGE), which incorporates the constraints of certain conserved quantities serving as integrals of motion. By drawing an analogy between eigenstate reduced density matrix and GGE, we conjecture that a natural set of conserved quantities for GGE can emerge from the reduced density matrices of properly chosen eigenstates by the entanglement Hamiltonian superdensity matrix (EHSM) framework, and we demonstrate this explicitly for models mappable to free fermions. The framework proposes that such conserved quantities are linear superpositions of eigenstate entanglement Hamiltonians of a larger auxiliary system, where the eigenstates are Fock states occupying what we call the common eigenmodes, which remain eigenmodes when truncated within the physical subsystem. For 1D homogeneous free fermions with (anti)periodic boundary conditions, which maps to 1D hardcore bosons with nearest neighbor hoppings, these conserved quantities lead to a non-Abelian GGE, which predicts the relaxation of both fermion and boson bilinears more accurately than the conventional Abelian GGE. Generalization of this framework may provide novel numerical insights for quantum integrability.

Music colleges in China currently face significant challenges in innovation and entrepreneurship education (IEE), including inadequate instructional support systems and fragmented teaching methods. These issues are particularly pronounced in aesthetic education courses, where scientific assessment and personalized guidance for students' entrepreneurial competence remain underdeveloped. To address this gap, this study proposes and validates an auxiliary instructional system based on the back propagation neural network (BPNN) model aimed at improving the precision and effectiveness of IEE within aesthetic education. The study investigates whether the BPNN model can accurately model and assess students' entrepreneurial abilities to support pedagogical optimization. Targeting graduates from music colleges in the Xi'an region, this study constructs a competence evaluation framework comprising four primary indicators and twelve secondary indicators, gathering 444 valid questionnaire responses. Using this data, the BPNN model is designed and trained to predict and provide feedback on students' innovation and entrepreneurship competencies. The model achieves a maximum relative error of only 1.64% between predicted and actual outputs, demonstrating strong accuracy and practical viability. Results highlight the theoretical and applied value of leveraging deep learning for entrepreneurial competence assessment in the integration of arts education and IEE. However, the current evaluation framework requires further refinement to better meet the evolving demands of specialization and industrialization.